Re: stefan's bittagging patch

Dan Coby wrote:

>
>Thanks for the information about the bit tagging device.  It looks like
>it will only be useful for a PCL type RGB device.  Is this is what 
>is desired?
>
>  
>
That was the intent when it was written, it was originally done as a 
demo for IBM.  I don't know if it will suffice for Teco.  The teco 
device requirements seem to be in a state of flux.  I don't know 
anything about the host based projects you are referring to.

As far as the device macros go customers shouldn't be allowed to touch 
that chinese finger puzzle.  The staff can't even use it effectively, 
how do you expect a customer to use it?  They do, on the other hand, 
understand a c structure layout.  I concede comments would be nicer.  
But this is out of my jurisdiction and not something for me to argue 
about.  You decide.

As far as the issues of correctness that Igor brought up.  The intent 
was to label charpath as vector not text and I answered the questions 
about cache and alpha in a previous mail.  I didn't look at the comment 
he reviewed but I am sure that he is probably right and it is misleading. 

As an aside, I am concerned that you are expending a great deal of 
effort on doing these tagging-like projects in the postscript language.  
I've already been down that path (doing it in the language).  It will 
not work for pcl.  I don't think a language based solution works for 
postscript either but I won't argue that point, I know it won't work for 
pcl, and thus your efforts will not be of use to teco.  If your solution 
involves only modifying the C implementation of the postscript operators 
in the graphics library I am sure whatever you get working with 
postscript will work fine with pcl, but I don't get the impression that 
is the approach you are using and I haven't seen any code.

My last suggestion about bittagging a few meetings back was to leave it 
in the pcl tree.  Y'all want a grand unified tree so you will need to 
deal with the consequences of that choice.

Henry